Efficacy and safety of antimicrobial de-escalation of treatment for sepsis A protocol for systematic review and meta-analysis

Background: Sepsis has become a global healthcare problem and continues to be one of the leading causes of death due to infection. In essence, early recognition and diagnosis of sepsis is needed to inhibit the transition into septic shock, which is correlated with higher mortality. Many studies have suggested antimicrobial de-escalation as one of the strategies to replace the empirical broad-spectrum antimicrobial treatment using a narrower antimicrobial therapy, especially among patients with sepsis. However, antimicrobial de-escalation therapeutic effects in sepsis remains unclear. We therefore performed the present study in an attempt to assess efficacy and safety of antimicrobial de-escalation therapy in patients with sepsis. Methods: We will carry out a systematic literature search to establish the potentially eligible trials from electronic databases, including EMBASE (1980 to October 16, 2020), MEDLINE via PubMed (1966 to October 16, 2020), Web of Science (1965 to October 16, 2020), Cochrane Library (CENTRAL; 2020, Issue 10), WanFang databases (last searched October 16, 2020), and China National Knowledge Infrastructure (CNKI; last searched October 16, 2020). For this study, the language restrictions are English or Chinese. Two authors independently examined quality based on the Cochrane Risk of Bias Tool V.2.0 and extracted data. Data obtained from the study will be synthesised using applicable statistical methods. Results: The results of the present study will systematically assess efficacy and safety of antimicrobial de-escalation therapy among patients with sepsis. Conclusion: The results of the present study will help to establish the efficacy and safety of antimicrobial de-escalation to treat patients with sepsis. It can also help to identify the most efficient and safe therapeutically-relevant method. Ethics and dissemination: The present study is a meta-analysis and the pooled results are based on published evidence. Therefore, ethics approval is not necessary.